|
Description
|
This dataset contains the supplementary document to the article 'Exploring the 3D Architecture of Rat Brain Tissue Using Digital Holographic Microscopy'.
Abstract:
Understanding the complexity of the (human) brain requires mapping its microscopic architecture to reveal neuronal cell structures and densely packed nerve fiber networks. This task traditionally relies on invasive methods, such as sectioning samples into thin slices and applying labeling techniques for microscopy, which can obscure or destroy valuable information and often demand extensive computational post-processing. A promising alternative is Digital Holographic Microscopy (DHM), which has been widely used for imaging transparent biological samples with minimal preparation and high resolution. DHM enables both phase and volumetric imaging from minimal data, yet its application in brain imaging remains largely unexplored. In this work, DHM is used to measure both the amplitude and phase of rat brain tissue. This complex-valued information, combined with digital filtering and holographic propagation, enhances 2D structural visualization and reveals volumetric features by propagating the acquired hologram and applying autofocusing criteria. This approach successfully resolves the three-dimensional arrangement of crossing fiber bundles, a challenge in many imaging applications. Finally, the technique is demonstrated to be scalable, allowing the measurement of full brain slices spanning several centimeters in a few minutes.
|
|
Notes
|
Instruction:
Download (and if necessary unpack) the files. Please find below an overview of the folder structure.
The dataset is structured in the following way:
- "Raw_images": Contains the raw images for the measured rat brain slice. The images are manually acquired in a zick-zack order starting at the top right going down towards the bottom left. The last number (*_.pgm) is the image id, locating the image in the sampling grid, the first number (frame*) )is the frame number, which identifies the hologram sampling method
0: "Up" Image
1: "Down" Image
2: "Gabor" Hologram
4X indicates the magnification of the lense.
- "Processed_images": Similar nomenclature to "Raw_images". Here, the first number indicates the position in the sampling grid. The corresponding modality is subsequently indicated by the name:
- "amp": Amplitude (absolute value of the DSB hologram)
- "phase": Phase (argument of the DSB hologram)
- "PC_x": digitally applied phase contrast image. x = 1,2,3 (applied retardation): 1 = 0.25 wavelengths, 2 = 0.5 wavelengths, 3 = 0.75 wavelengths
- "DF": digitally applied dark field image
Subfolder "refractive_index" contains the refractive index maps shown in the paper in Figure 4.
- "Propagation": Folder structure for single regions indicated by sampling grid positions 34, 62, 103, 127, and 132 as described in the paper.
- Folder "amp" contains the amplitude and phase contrast with a retardation of 0.25 lambda ("PC"). Folders contain propagated images with incrementing integer enumeration in z-direction, i.e., .
"1" corresponds to a propagation distance of -100 microns,
"101" is the initial image and
"201" corresponds to a propagation distance of +100 microns.
- Folder "videos" contains a scrolling animation of the propagation distance where the current axial position is indicated in the video.
- Folder "autofocussing" contains the propagated images of the region "34" which are evaluated with the Tamura coefficient of the gradient image as depicted in Figure 5 in the paper.
- "Simulation": Contains the results for the simulation of the DSB hologram acquisition as matlab figures with the following nomenclature:
- "t_*": amplitude scaling from 0 to 1;
- "p_*": phase scaling from 0 to 2 pi;
- "DSB": initial DSB hologram (eq. 5)
- "DSB_opt": optimized DSB hologram (eq. 6)
- "Hilbert": Different contributions of the Hilbert transforms of R in eq 5
- "input": scaled input images that define the measured light field.
|
v. 4.20 build v4.20-with-fzj-features
